Presentation Ninja

# Quantification of Behaviour in Neuroscience

### Jennet Baumbach

### September 15th 2023

---

## The Importance of Behaviour to Neuroscience

### **Behaviour**: The activity of an organism interacting with its environment.

- Naturalistic behavioural patterns are not synonymous with stimulus responses.

- In laboratory settings, we primarilly study stimulus responses.

### **Monistic** view point: behaviour is the summated output of the underlying neural processes.

- Alterations in behavioural output represents changes in the underlying neuronal processes

---

# Historical Perspective

### The mind-body problem (aka monism vs. dualism)

--
    
- **Dualism**: the idea that the mind and body are separate (distinct) entities

- Aristotle (300 BCE), Plato (400 BCE), Rene Decartes (1600's)

- Rationalized that the body and the "soul" or the "mind" were clearly distinct

--
    + the death of the body is separate from the existence of the soul

---

# Historical Perspective

### Early neuroscience (1800's - 1950's) relied heavilly on behaviour.

Clinical case studies: e.g. [Phineas Gage (1848)](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7735047/#:~:text=Gage's%20case%20is%20considered%20to,personality%2C%20emotions%20and%20social%20interaction.&text=Prior%20to%20this%20case%2C%20the,and%20unrelated%20to%20human%20behavior.)

.pull-left[
<img src="data:image/png;base64,#https://upload.wikimedia.org/wikipedia/commons/2/28/Phineas_gage_-_1868_skull_diagram.jpg",width="350", height="300"/>
]

.pull-right[
<img src="data:image/png;base64,#https://upload.wikimedia.org/wikipedia/commons/thumb/1/1b/Phineas_Gage_Cased_Daguerreotype_WilgusPhoto2008-12-19_EnhancedRetouched_Color.jpg/600px-Phineas_Gage_Cased_Daguerreotype_WilgusPhoto2008-12-19_EnhancedRetouched_Color.jpg",width="350", height="300"/>
]

**Post-injury behavioural patterns** provided support for cerebral localization and information about the role of the damaged brain region (PFC).

---

# Historical Perspective

- The dualist perspective was challenged by increasing support for the connection between brain and mind during the 1800's

--
    + Rather than dualism being "disproven", science adopted the monistic viewpoint because of the volume of evidence to support it.

- [Paul Broca (1861)](https://en.wikipedia.org/wiki/Paul_Broca) linked behavioural defects in clinical populations with brain lesion (identified during post-morterm analysis)

--
    + Indicated that there was localization of specific cognitive function distributed across the cortex.
    
--

<img src = "https://i0.wp.com/biomedguide.com/wp-content/uploads/2019/10/Broca_Wernicke.jpg?fit=487%2C301&ssl=1">
    
---

# Historical Perspective

### 1900's-recently, focus shifted towards molecular approaches.E.g.:

.pull-left[

<img src = "data:image/png;base64,#https://neurosciencenews.com/files/2022/06/learning-memory-mechanism-neuroscience.jpg" height = "250", width="400"/>
]

.pull-right[

- A complex behaviour (here, freezing behaviour) is reduced to a mechanisms represented at the level of a single neuron.

- Has provided many pieces of insight about the molecular processes that contribute to to complex behaviour

- Fundmentally lacks qualitative information about *how* these processes produce the behaviour in the context of the system.

]
---

# Historical Perspective

### Recent renewed interest in the study of behaviour

- Behaviuoral work can aim to *understand* in ways that neural interventions cannot. 
    +  Complex behavioural characterization should precede the formation of neural hypotheses.

- Criticism of approaches that involve inference of cognitive processes from neural processes (e.g. [mirror neurons: Caslie et al. 2011](https://www.semanticscholar.org/paper/The-Mirror-Neuron-System-Casile-Caggiano/a3357d5d183cdd47ecbe352f34d7aec7e37fd8a0))

--
    + And clear separation of results versus interpretations. 
    
--

.footnote[[Neuroscience needs behaviour: correcting a Reductionist Bias (Krakauer et al., 2017)](https://www.cell.com/action/showPdf?pii=S0896-6273%2816%2931040-6)]
    
---

# The Connection between Brain & Behaviour

--
    
<center>

</center>

The **result** is that similar patterns of neural firing were observed when the individual performs the behaviour, or, when the individual watches another perform that same behaviour.

The **interpretation** has been that the patterns of firing represent some levels of "understanding".

---

# Historical Perspective

### Recent renewed interest in the study of behaviour 
  
--

- Technological developments improve the speed and accuracy in collecting behavioural data.

- Producing more complex datasets

- Which facilitate more comprehensive analytic approaches

---

## Animal Behaviour as a Readout for Brain Function

### Lesion studies

- A loss of a specific behaviour after a specific brain region allows researchers to infer the functionality of the brain region.

</center>

.footnote[[Hutcheson et al. 2022](https://link.springer.com/article/10.1007/s002130000635)]

---

## Animal Behaviour as a Readout for Brain Function

### Lesion studies

- A loss of a specific behaviour after a specific brain region allows researchers to infer the functionality of the brain region.

.pull-left[

<img src="data:image/png;base64,#latency.png" width="815" />
]

.pull-right[

]

<b>The lesion approach does not inform about <i>how</i> the brain region contributes to the behaviour</b>.

</center>

.footnote[[Hutcheson et al. 2022](https://link.springer.com/article/10.1007/s002130000635)]
  
---

## Animal Behaviour as a Readout for Brain Function

### Developmental perspective

- Changes in behavioural output across the lifespan suggest alterations in brain maturation.

</center>

.footnote[[Shoji et al. 2016](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4730600/)]

---

# Behavioural Paradigms in Research

### The importance of well-defined behavioural tests

#### Reliability (*does the measure produce consistent results*)

- test-retest reliability (within or between individuals)

- inter-rater reliability

#### Validity (*are you really quantifying what you think you're measuring?*)

- face validity - Does it "check out"?

- pharmacological validity - Do pharmaceutical agenst that are known to regulate the processes that the measure is assumed to capture alter the behaviour in the expected ways? 
  + e.g. administration of an anti-anxiety drug should increase anxiolytic behaviour..
  
--

- translational validity

---

# Behavioural Paradigms in Research

### Example: Tasks to investigate memory in the lab

Non-aversive memory measures

- Operate on the basis that animals have a **preference for novelty**

+ Robust cross-species examples of this preference

---

# Behavioural Paradigms in Research

### Example: Tasks to investigate memory in the lab

Aversive memory measures

- Operate on the basis that animals fear pain

+ Robust cross-species examples of this behaviour 
  
--
  
<center>

---

## Tapestry Analogy

- Courtesy of [Prof. Dawn Good](https://brocku.ca/social-sciences/psychology/people/dawn-good/)

.pull-left[
<img src="data:image/png;base64,#https://previews.123rf.com/images/niknikpo/niknikpo1503/niknikpo150300010/38075069-colorful-gobelin-tapestry-texture-close-up-background.jpg",height="250",width="250"/>
]

.pull-right[
<img src="data:image/png;base64,#https://www.metmuseum.org/-/media/images/blogs/now-at-the-met/2014/making-a-tapestry/boreas-and-orithyia.jpg",height="250",width="250"/>
]

---

## Challenges and Limitations

### Variability in animal behaviour

- Both between and within individuals

### Complex Nature of Animal Behavior

- Contextual considerations: generally animals are tested one at a time in novel behavioural contexts.

- Natural complexities: very difficult to confidently identify playing versus aggressive behaviour

### Animal Origens

- Selective breeding for desireable lab characteristics has unknown influences on animals' behavioural output.

---

# Hierarchical Framework of Scientific Thought

### **Theory**: A unifying scientific concept that is supported by many individual pieces of evidence.

- Examples: Evolution, relativity, synaptic plasticity in neuroscience.

- Individual scientists do not generally interact directly with theories on a day-to-day basis

--
    + e.g. no grant proposal aims to tackle a theory.

- Acts as the conceptual starting point for researchers.

---

# Hierarchical Framework of Scientific Thought

### **Model**: deals with one chunk of a scientific theory

#### (1) Unifying models

Are large, and typically fairly crystallized.

- e.g. the genetic basis of inheritance, epigenetics, LTP in neuroscience. 
    + Provide mechanistic insight into the overarching theory of evolution.

#### (2) Working models

Which are typically smaller in scope and often more flexible.

- A unifying model could encompass many working models

- e.g. Epigenetics: DNA methylation, histone modifications. Neuroscience: changes in dendritic shape / density .

---

# Hierarchical Framework of Scientific Thought

### **Hypothesis**: isolates one experimentally-testable aspect of a a working model.

- Facilitates the formation of *testable predictions* that are usually directional in nature.

- A hypothesis can never be "proven correct" - it can only be "not disproven"

--
    
- A hypothesis gains support through repeated testing of its predictions.

- The wording of the hypothesis should lend itself well to discussion and rigorous experimental design.

- "Hypothesis driven" writing is the gold standard for research reports, grant applications and thesis projects.

- e.g. in LTP, could form a hypothesis: synaptic strengthening facilitates long-term memory via the NMDA receptor

---

# Hierarchical Framework of Scientific Thought

### **Experiment**: An emperically designed series of steps designed to investigate the predictions that arise from the hypothesis.

- Involves some kind of control group.

- Either quasi-experimental or true experimental design.

--
    + quasi- examples: Males versus females, clinical populations versus healthy control group.

- Differentiation from "study" in psychology.

---

# Hierarchical Framework of Scientific Thought

### **Measure** or **Assay**: A specific endpoint that will produce interpretable data.

- e.g. a specific behavioural task or laboratory assay.

---

# Styles of Experiments

#### 1. Observe: An initial observed association between two things (e.g. A, C) initiates the research design process.

Design hypothesis: A -> B -> C

#### 2. Measure: Confirm that A causes B

#### 3. Block: inhibiting B should prevent A from causing C

Design an experiment to prevent B while leaving the processes that cause A and C intact (necessity)

#### 4. Mimic: experimentally manipulating B should prdoduce C

Design an experiment to alter B to investigate whether this will produces C in the absence of A (sufficieny)

.footnote[[David Sweatt - Mechanisms of memory textbook ](https://www-sciencedirect-com.myaccess.library.utoronto.ca/science/article/pii/B9780123749512000044?via%3Dihub)]

---

## Future Directions in animal behavioural research

#### Improvements in the quantification of behavioural data provide richer, more complex datasets.

- Much more efficient (e.g. eliminiates the need to sit scoring videos for many hours)

- Can use an exhaustive approach combined with unbiased clustering methods.

.pull-left[
![](data:image/png;base64,#https://images.squarespace-cdn.com/content/v1/57f6d51c9f74566f55ecf271/1665060903598-UK6UI28ND04D6OPLXKSG/ezgif.com-gif-maker%2B%289%29.gif?format=500w)
]

.pull-right[
![](data:image/png;base64,#https://images.squarespace-cdn.com/content/v1/57f6d51c9f74566f55ecf271/1548190683014-WE15RBTQKNB3DZFE6RSY/R0226_20180110_15-35-01_030_direct.gif?format=500w)
]

---

## Future Directions in animal behavioural research

### Improvements in the complexity of statistical analysis allow for richer interpretatins about behavioural patterns

- e.g. Time-series based analyses (a concept borrowed from finance)

---

# Questions ? :)

---